we climb the mountain or do we stay at home and go out in the
afternoon to see a film. Climbing the mountain, one of us may slip
and be killed. We may be killed walking to the cinema, but there’s
less chance of it. Climbing is not very dangerous, but it’s danger-
ous enough for us always to prefer an alternative way of spending
our free time, so long as we reason in maximin fashion. Were we
maximiners we would never venture onto a steep slope. There is
always something safer we could be doing.
What is the alternative to maximin? It is time to bring the rea-
soning in favour of (2) out of the woodwork. As (2) suggests, it is
maximum average utility. Between classes A and B, supposing
members of these classes are equal in numbers, one might suppose
that some people, those who like a gamble, would compute average
expected utility at 90 units – and go for it. They may find them-
selves in the class who receive 30 units, and worse off than they
would be under maximin, but they may be better off than they
would be under either equality or the difference principle. If we do
the sums we find that average utility under (2) will amount to 90
units (30 plus 150 = 180; the sum divided by 2, = 90 units of util-
ity). Computing in the same fashion, the average utility of (1) is 50;
the average utility of (3) is 60 units of the primary goods. Since the
utility of (2) is greater than the utility of either (1) or (3) why not
go for it? The objection to Rawls can be phrased more strongly as:
what reasons are there for not taking the approach of average
utility, gambling on the chance of being one of the better off, gain-
ing 150 units, and risking the prospect of losing – receiving 30
rather than 50 under equality or 55 under the difference principle?
Rawls’s answer is that we wouldn’t dare.^63 We only have one life
to lead and the basic structure of the society in which we live is
crucial to our well-being, and just as importantly, to that of our
children. We would be wrong to risk the possibility of receiving 30
units when we can guarantee the receipt of 50 units or better. The
utilitarian, as ever, has a cogent reply. In the comparison of (1) and
(2), if (2) is represented as an outcome that the proponent of max-
imum average utilitarianism would endorse, either like is not being
compared with like or the situation is underdescribed. It looks as
though like is not being compared with like since the utilitarian
will be concerned to envisage outcomes primarily in terms of the
distribution of utilities, rather than primary goods. As we saw in
DISTRIBUTIVE JUSTICE